Leveraging Customer Data Using CDP to Build a Great Personalization Strategy

Between the two of them, Bas and Joost have close to a decade of experience, implementing personalization for fast-growing companies. And in this session, they are going to share some nuggets from their experience, implementing those personalization strategies for those companies. They are going to share a practical roadmap that you and I could implement to deploy personalization on our digital properties. If you have a website, web app, or anything in between, this session would be a good primer to understand how you can implement personalization. 

Bas and Joost, thank you so much for giving us the opportunity to host you. And, before I give the mic over to you, I’m gonna turn my camera off so that the focus stays on your folks only.

Joost Kuijlaars:

Thank you, Ajit. Good afternoon, everybody. We are honored that so many people are attending this webinar. My name is Joost Kuijlaars. I’m the founder of The Data Actors, a service and consultancy agency in the field of data-driven business. I started in the nineties in the data area in the middle of the company. At that time, it was shopping by catalog. And I have been in the data business since then for several other branches, like retail, eCommerce, B2B, publishing, and fundraising. So data is in my genes. In this webinar, I want to convey my experience and knowledge on how to make data actionable, especially on the personalization subject.

Bas Cuperus:

Thank you very much, Joost, for introducing yourself, and thank you very much, Ajit, for your, for your words. Good afternoon to you all. My name is Bas Cuperus. Unlike Joost, I did not set my first steps in data-driven marketing in the 90s. But just my, first steps. I started my professional career a bit later, namely in 2017, and for most of that period, I’ve been working as a consultant, for Invenna. So at Invenna, we deliver software solutions for marketing, and customer data platforms which we will talk about a little bit more during this webinar, we deliver solutions for IT, and compliance in KYC and transaction monitoring solutions. Well, I’m mostly working on these, customer data platform implementations marketing automation projects, yeah, mostly adding data sources, channels, and and, all kinds of connections with the relevant applications. So, basically, I’m helping our customers to get the most out of their 360-degree customer view.

And, it’s an important condition for personalization. So, my goal for today is, like, twofold. So first of all, I would like to just show you how you can combine your online and offline data. To enable you to do personalization on online channels based on the data you collected from both online and offline sources. My second goal for today is to show you why a 360-degree customer review is important for personalization.

So well, I was preparing for this webinar. I was searching for some personalization efforts. So it would be better off use that complete 360-degree customer view, and, of course, corresponding business rules to form a clear uniform message. 

So I took 2 examples. The first one is, from a workshop. They sell many different things from merchandising to clothing and different music. It’s a good personalization effort. It’s an effort to reduce churn. They also use a product for a recommender at the bottom. And the churn effort is a promotional code, which gives me €10 discount on my next, spent.

I only buy records at this store, and the only thing I bought, and you can also see it in the recommender, in the product recommender. And, the set thing is that this code is not valid on recommended products. They don’t use all the information to personalize this, properly. 

Well, the second example is, is for like. It’s a retention email as a reminder for my Barber. Let’s say, like, it’s been, 5 weeks since you’ve been here. Maybe you should come again, which is a great effort, of course. But I already made a new appointment. So this causes a little bit of confusion. It’s also not using all the information you have to personalize your, effort, which is a shame and, that’s something we are talking about today.

Joost:

Interesting cases, boss. I think your barbershop has probably not connected his appointment system. But more importantly, in both cases, they are busy with some kind of personalization. Think you will get here the first poll for today. Please fill in, and I will continue with our presentation.

We learned that personalization is the number one priority for more than 50% of all marketers. Well, Gartner says that they expect that 80% will stop with personalization in 2025 due to 2 reasons. First of all, the insufficient data management, we saw in both cases. Second of all, the lack of return on investment. In this webinar, we will elaborate those both so you can become successful in personalization. 

Let us briefly take you to the agenda for today, starting with the definition of personalization. Then following the subject on how to know your customer and using the customer data platform. Then interpret the data paradox in relationship to personalization, Bas will take us through the tech manager’s use for web personalization, and finally, we will end with the learning cycle. 

Let’s start with the definition of personalization. The first part is, of course, about the person. Marketers are used to working on averages. But if you take the average Dutch guy, for example, boss and I, you will get a guy with 1 brown and 1 blue eye. The average person does not exist in personalization. Each person has their own characteristics, interest, and motivation to do business with you, for you to find the right one. 

The second part of the definition is about the moment. It’s not about the right time on the clock. It’s related to the time in a customer journey or sales funnel. Is your customer just orientating or perhaps ready to buy or even just bought an item in your store, the message you will give him will differ per moment. The use of the channel is part 3 in personalization. Although we have an omnichannel customer, she or he will always have a preference to communicate with you or to do business with you.

But not letting go, it’s the only channel, but it is still one of the biggest challenges of most companies, and it’s organized in silos. It is important to make the customer experience consistent over the channels. And finally, to complete the definition, it’s all about content. Content depends on the stage of the customer journey, on the customer’s interest, or even his behavior at your company. This content has to give value to the customer. Like the recommendations in the example from us, which will finally give also value for you. 

And then we have the definition of personalization. Give the right customer at the right moment, by the right channels, the right content.

Bas:

Absolutely. So in order to personalize and to make use of the right person, the right moment, the right channel, and the right content, we need information. So in this section, I want to elaborate on the role a customer data platform can play to extract and structure this information that we need in order to be successful in personalization. 

So we start with, what is a customer data platform. How do we define it? 

The Custom Data Platform Institute stays at CDP is packaged software that creates a persistent, unified customer database, that is accessible to all other systems. Well, we can break this down into three elements, like the packaged software, the customer database, and the accessibility, and I will discuss, like, the latest two. 

So first of all, we’re talking about the persistent unified customer database

What is it? Well, the customer data platform creates a comprehensive view of each customer by capturing data from multiple systems. As you can see on the left hand in this figure, there are, yeah, lots of different data points, which capture customer data. Across, a lot of different systems. We’ve got 1st party data in our own landscape, which, is maybe a CRM system with personal and demographic data, data from the socials and research, from transaction and system agreements, and contracts, and maybe indeed online behavior. So a lot of different, yeah, sources and a lot of different customer data, which maybe we can add some reference data from, from a second part to show movements, maybe some segment data from second parties.

We have all this data over multiple systems, and we try to link that information related to the same customer and store that information to track behavior over time. So there’s a part of data integration. We validate this data, we clean all of these, points and we do some kind of identity resolution. Are we talking about the same customer? Is it a different customer? All these data sources should be combined to see the unique customer. When we have that unique customer, we build a 360-degree view around him on which we can perform, insights and analytics maybe some recommendations, even predictive modeling, and, of course, also contain, management and marketing automation, which is, important for the personalization we are talking about. 

Well, the figure on the far right is the accessibility to all other systems. So the data we store in the 360-degree view can be used by other systems for analysis and to manage customer interactions so that we can personalize on, the website, for example, but also in our emails, in our DMs, etc. 

So, yeah, that 360-degree view, that that’s necessary. Because in the end, we have a 360-degree customer view, which provides us the information on who is the right person, and what his or her behavior looks like. And what are the products he is interested in? It provides us, information on the right moments where it’s our customer in the customer journey, where is he in the funnel, but also on, yeah, what what timings is there, the most interaction during the evening or maybe, during the weekends or working hours, same for the right channel. On what channel do we have the most interaction with the customer which channels are used for gathering information by our customers, and which one, is for the purchase? So an example of that 360-degree view, you can you can see here right now. So we combined information from lots of different quarters. We see some website visits. The purchase comes from different sources. And it is united in one 360-degree view so that you can personalize all available information.

So I was wondering, I think we have the second poll, right now. Yeah. Do you have, a 360-degree customer view, or, are you trying to build one?

Joost:

Thanks, boss. Ah, Pat Johnson is a really rich profile. You can make a good personal case personalization case on it, but that’s not certainly for every customer. Let us take a simple model where Pareto still rules. Just a small part of all your customers are responsible for the main part of your revenue or completely for your profit.

Is it 20-80, 25-75, or 30-70? It’s a small group of your total customer base. The mid group will only mostly provide revenue and the bottom group costs you a lot of money. On this pyramid, there are 3 main marketing strategies applicable – lead generation or prospecting on the bottom side, improving the lifetime value, and, retention or preventing churn.

For the data, you will find the reverse pyramid, which is logical, and creates the data paradox. A small part of your customers will generate the most data because they have the most contact with you in your newsletter or your website with your contact center. With the majority group at the bottom of the pyramid, you will have hardly any contact for just a small data set. For personalization, you really need the data. A lot of data to find patterns or to build AI models. These models or patterns will enable you to personalize and apply automation to retrieve a return on investment.

Related to the main strategies, personalization is limited to the top of the pyramid.

Bas:

Thank you, Joost. Yeah. My second goal for today, if you can recall it, was to show you how you can combine your online and offline data. To enable you to do personalization, online channels are based on the data you collected from both online and offline sources. So you’re supposed to talk about the data paradox. For a large group of customers and prospects, we have very little data. I would like to show you now from a, yeah, not really a technical point of view, but, a bit technical how to combine, like, those online data, with the data from all your offline sources and, mostly how to boost this, this combination. I will focus, on the incoming data from from your website or workshop. And I will also show you how you can use a customer data platform, tag manager, and other combinations. To eventually do personalization on your website.

So let me first show you how to get online data into the custom data platform and mess it with the right offline customer. So, between the website and the customer debt platform, in the figure, you now see that a tech management system, appears and I would like to show you in this story how it helps us, to combine the online and offline data.

Can we get the next slide here? Thank you. So what do we do with the tech manager? So, essentially, we track website behavior on the website and store this behavior in different values. So, the values can, can be a visitor, and events where the visitors are the unique, yeah, preferably first party cookie that refers to a device, which essentially, is a person.

It’s not 100% but that’s, that’s the thing, of course, with the online data. The second one is events. It can be anything that you would like to track, from a page visit to a form submission and I can scroll that on your page to see whether someone read the article, you provided on your page or not. 

The third one is like a session, which is, a key that groups all the events in one session. So if you make the right text, we can do analyses that will help to answer questions related to the right person, right moment, right contents, etc. And, there’s a root from the tech manager to the customer data platform we extract those, data points and load them into the customer data platform. The other way around, we can also use the data in the customer data platform to sense certain values to the tech manager. From that, we can write off again, first-party cookie values on which you can personalize. A few days later, I will show exactly how this works, but for now, it’s good to remember that the tech manager delivers data to the customer data platform. But the customer dates plus will also deliver data to the tech manager.

So back to the question, how do we match that online data from the tag manager with the customer data platform data coming from all offline sources? So, essentially, we can do this in two ways, and the first one is quite obvious most, most websites. A lot of websites use the possibility to, with a login So if someone logs in, we obviously know who visits our websites, and we just, yeah, make to, have to put up a right date stream that provides the needed data to the customer data platform. So with a log-in, it’s rather simple. But most people, of course, do not log in or are not logged in during the session. 

So, there’s another route that we can, explore, and the second route is via clicks in, for example, email messages. Well, there we make use of parameters. Probably you already use them in, email marketing, UTM tracking is one of the most famous, parameters here, and it’s based on the, yeah, exactly the same principles. So let me explain this in the next slide. 

So today, we all received an email from VWO with a link to attend this webinar. Well, that email does not have very much clickable links to the VWO website. So we only have to focus now on the joining link. And the joining link looks like if you press the next host, then we can see the thank you. To that number that we see in that URL, is probably a unique combination between me as an attendee and a specific webinar. This way, VWO can allow my behavior during the webinar.

Did I attend the full webinar? Did I interact during the webinar? What triggers for me what triggers work for me to interact? And with this information, VWO can start with personalizing messages to me. What is the next interesting webinar based on my earlier tendencies and what type of triggers, VWO can use, to let me engage more during the webinar? What goal to actions work, best for me, etcetera. So What we can do in our email messages is to add a parameter that represents me. So if VWO now made a unique link, we can add a parameter to the URL to also make it unique. 

The tech manager can get the parameter from the URL and send it to the customer data platform. The customer data platform knows that that number corresponds with me, and so it can enrich my customer profile with the web data from that session after a click. The last step is to make use of the data stored in the customer data platform to personalize the website. So we can make personalization, values, like, the next best action or next best offer in our customer data platform, based on all the information we have on the customer, and send them to the tech manager. 

The tag manager stores these values in a first-party cookie. And maybe you can click a few tons more on the slide. Yeah. Exactly. So the tag manager stores this value in a first-party cookie and VWO. Can read out a cookie and based on the value, personalize the web page. So, in our customer data platform, we know that my email address corresponds to a certain ID, which we added as a parameter in your, URL. 

If, yeah, someone clicks on that email link, we’ve got now the combination between the visitor ID from the tech manager and the parameter we created in, in our customer data platform. So we know this person corresponds to this web ID and this visitor ID, and, yeah, based on if someone comes in with that web ID or with that visitor ID, we can write a first-party cookie value, for example, a string with, churn risk high. The next best offer and newsletter subscriber is zero. So based on those values in that string, we can build the logic and VWO to eventually personalize, the web page. 

My second goal was to show you how to combine your online and offline data, to enable you to do personalization on your website. This is how that essentially, works in most cases. Next up, Joost is going to tell you about how to make that data actionable, and let’s talk about the learning cycle.

Joost:

Greg Boss, technically, everybody everything’s possible, to succeed in, ROI-driven personalization, you have to realize it’s a long-term process of connecting data to execution. Or as we define it, making data actionable. First, you have to collect all connect contact data with your customers and connect to a single customer view, but show us how. Focus on the contact points, which are most likely to be used, and hard drive, conversion, and revenue. The second part of this learning cycle is to analyze the data and create insights to find patterns by human interpretation or by using AI.

But at the same moment, you have to challenge your marketeer to use his gut feeling. And guts to find potential use cases, which improve the customer experience and, therefore, increase your revenue. Increase your revenue. 

The next step logically is to test the ideas and look for confirmation or the interpretation of the insights. Run the test in the most relevant touchpoints for AB testing, and where marketing automation can be used, and do not forget to prove that the test generates an uplift in return on investment.

Finally, if you have a successful test, you have to scale it up, the test to other relevant segments, channels, and touchpoints, ensuring your customer the best and consistent experience across all channels. At this moment, you will create your flying wheel. To improve your sales, obviously, and really improve your return on investment. So give me a simple example of one of our cases. For kitchen equipment, we added the first step and made the data for products.

In the second step, analyze the win-back campaign for the former talk customers. This campaign was losing its effect on the churn. We discovered that for a substantive part of the churn, we didn’t buy products of a specific brand. In the last months, no product of the brand was offered in win back. So in our Windback case, email was the most effective channel, and we decided to do 2 tests, one by adding just one product of the brand from the portfolio.

And besides a test where the email was more in look and feel of the brand. A single product test yielded a plus of 10% and the look and filter have used an uplift of over 20% and were seeking significant improvement of the ROI. For our kitchen equipment, the look and feel campaign was also applied to the landing page. And finally, the whole approach improved the ROI of this group by more than 50%. So you see that, the return on investment is improving, not just doing the test, but also, to, extract that test to the other channels and other contact points at the customer journey.

Bas:

Great. Impressive, numbers, Joost. Yeah. The key takeaways for today. I tried to make clear, at least, that that like, the full body of information, the 300 and 60, the green view is necessary for personalization.

So that you have information on, who’s the right person, what’s the right moment, what is the right channel, So when we can remember the examples, which are good efforts in, in personalization, but they don’t use the full potential of, of all the information, scattered around the, the organization.

Joost:

Yes. The second, second takeaway we will give you is, to analyze and focus on the most effective data points. And execution. Realize it’s a long-term process with successes.

Bas:

Absolutely. And, yeah, well, my first key takeaway was to make use of parameters. So, like, remember the web ID, and super parameter in the URL in every moment of contact. It gives you the possibility to match your online data with your offline data, and that is important to enrich, 1st of all, the customer profile with website behavior, and interest, and it gives you secondly, the opportunity to personalize web pages.

Joost:

And the final takeaway for today. We didn’t discuss it literally, but it’s essential. Personalization is a process that concerns all your departments, not only data, not only marketing online or email. It’s cooperation together to get the best results. That’s also on our part, this webinar, and do we have questions?

Ajit:

Yes. We do. Thank you, Houston and that was an interesting presentation I would also have to congratulate you that out of you know, all the people who joined during the beginning from the beginning of the webinar, I see that 80 to 90 percent of people have still stayed at this point. Okay. So, that’s a massive number.

Normally, we don’t tend to see that many people learn because people get different ideas. They get hungry, or they go to the kitchen to get water, and they end up staying there only. So, you know, this is a a great great number that I see. You have managed to engage until the end and I love it. Now, with regard to the question, we have got one question from Peter and he’s asked, how do we find customer DNA?

So I think at one point, I particularly love that question and I would also have the same question. At one point, I think you guys talk about, customer DNA. How do we find customer DNA? So do you have any recommendations there? As a company, how can I effectively find my customer’s DNA?

Bas:

Yeah. Of course. First of all, I would say just combine all the sources that are, that have yeah, customer data in them to a central point, so make the customer your focus and then, yeah, and model your data around that customer.

So when you combine all those sources, that we discussed in the presentation, we’ve got quite a clear view of, who’s the customer, what drives them, etc. Yeah, mostly combine those in, all those sources, to find the relevant information. Okay. Maybe Joost can say something from an organizational perspective, on that.

Joost:

No. What I can say is that you don’t have the customer’s DNA all at once. It’s also a process where you start gathering insights and improving those insights, as a sole kind of learning process.

Ajit:

Oh, let me clarify this. When, you say insights, are you referring to quantitative data that we can gather from all these analytics tools and and and and and marketing stacks? When we say insights, could you, you know, give us, shine some light on it? Maybe explain it a little further.

Joost:

Okay. Can you repeat that, please?

Ajit:

Yeah. So when you say insights, are you referring to just quantitative insights or the quantitative insights that we get from the marketing stack like analytics or our email platform and so on? But What more? Uh-uh-uh uh-uh what more are you referring to when you say insights?

Like, how can we gather more insights for customers is what I’m asking?

Joost:

And now the way you get more, it’s about what was described, but, what I was saying is it’s more for your market, to understand what the target groups are. So you start with social demographics, if you have them, and finally, all content data, you will have a budget customer. So it’s not in detail, but you have to use, a little bit, of abstracted data from that. So, for example, we use, for several companies we are using, taxonomies. And text messages are related to the way they can provide, added value for the customer and added value for the company.

Ajit:

Okay. Right. Yeah. And also, you know, at VWO together, apart from the quantitative side, the other, medium that we use internally to gather a quality to pay our customer which could be used for a lot of people is session recording, heat map, and so on.

Do you guys have experience using these tools to gather insights for customers?

Bas:

You mean, heat maps on your on your website. So where do the people click the most? Well, it’s an interesting one. I haven’t used them, yet, because, yeah, it’s mostly to see how your website performs. But, yeah, that could also be quite useful, maybe as customer data.

Yeah, though we have to say, of course, all the events for the more you can trigger. I don’t know whether but I really have to think about the added value of a heat map, to all that, and other events and see whether that’s useful, for further personalization. But, yeah, it’s good to consider it, I think.

Joost:

Yeah. I agree with that, boss. I think it has to do with building up your DNA, your customer’s DNA. And, if at some value for you, improving your return on investment.

Ajit:

Great. So I think, that’s the question that we had, I don’t see any other question. I’ll still give, 30 seconds more. Okay. 60 seconds more to our audience.

If you have any questions now, it’s time to ask. If you don’t ask any questions, we’ll assume that you understood this presentation so much that you know, there’s no question at all. So, now is the time to ask questions, and yeah, the best question is going to get a free pizza from us. Okay. We’ll deliver it to your address. So, we’ll give 30 seconds more. 

No problem. You guys know how to reach out to, use them, and pass. The email address is on the screen. You can record it and you can reach out to them later on after this presentation. Finally, thank you so much for joining this webinar, and, thank you for the final session. I appreciate it.

Appreciate your time, and everyone, who is part of the session right now. And, thank you from here to Blue. All the best. Cheers.

Bas:

Yeah. Thanks a lot for your attention, and, and thank you very much, for moderating this session, Ajit. And thank you, Joost, for also sharing your knowledge.

Joost:

Thank you, boss, for the cooperation. Thank you, Ajit, and have a nice after afternoon. Bye bye.